[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20201207091312.cyi6rkt3owlkwsew@vireshk-i7>
Date: Mon, 7 Dec 2020 14:43:13 +0530
From: Viresh Kumar <viresh.kumar@...aro.org>
To: Valentin Schneider <valentin.schneider@....com>
Cc: Ingo Molnar <mingo@...hat.com>,
Peter Zijlstra <peterz@...radead.org>,
Vincent Guittot <vincent.guittot@...aro.org>,
Amit Daniel Kachhap <amit.kachhap@...il.com>,
Daniel Lezcano <daniel.lezcano@...aro.org>,
Javi Merino <javi.merino@...nel.org>,
Zhang Rui <rui.zhang@...el.com>,
Amit Kucheria <amitk@...nel.org>, linux-kernel@...r.kernel.org,
Quentin Perret <qperret@...gle.com>,
Lukasz Luba <lukasz.luba@....com>, linux-pm@...r.kernel.org
Subject: Re: [PATCH V3 2/2] thermal: cpufreq_cooling: Reuse sched_cpu_util()
for SMP platforms
Hi Valentin,
On 01-12-20, 17:25, Valentin Schneider wrote:
> PELT time-scaling can make the util signals behave strangely from an
> external PoV. For instance, on a big.LITTLE system, the rq util of a LITTLE
> CPU may suddenly drop if it was stuck on a too-low OPP for some time and
> eventually reached the "right" OPP (i.e. got idle time).
>
> Also, as Peter pointed out in [1], task migrations can easily confuse an
> external observer that considers util to be "an image of the recent past".
I agree with what you wrote and such issues may happen here as they
can in case of schedutil as well. The idea behind this patchset was to
get the allocator (IPA) and consumer (schedutil) in sync with respect
to frequency and power. It is better to allocate the power that
schedutil is going to request, then to allocate something based on
different metrics. If there is a problem with PELT signal then I will
let both the entities suffer with that.
--
viresh
Powered by blists - more mailing lists